Trial Outcomes & Findings for Persuasion in Medicine: Experimental Evidence on Sender and Signal Effects (NCT NCT04160975)

Last Updated: 2021-06-29

Results Overview

This is an inverse-covariance-weighted index comprised of responses to survey questions regarding: 1. whether the respondent was interested in further medical advice from the given sender (measured as a binary outcome); 2. whether the respondent trusted advice from the sender measured on a scale of 1 to 5; 3. and whether the respondent's assessment of the sender's qualification to provide medical advice measured on a scale of 1 to 5. It is standardized to a mean of 0 and standard deviation of 1. Scores less than 0 indicate lower ratings of the sender rated by participants in Arm A (described in each arm/group description) compared to the rating rated by participants in Arm B (also described in each arm/group description). Scores greater than 0 indicate higher ratings of the sender by Arm A compared to by Arm B. For example, an index score of .18 indicates that participants in Arm A rate the sender 0.18 index (in standard deviation units) higher than those in Arm B.

Recruitment status

COMPLETED

Study phase

Target enrollment

3245 participants

Primary outcome timeframe

This outcome was assessed during Baseline survey, which took approximately 20 minutes.

Results posted on

2021-06-29

Participant Flow

Participant milestones

Participant milestones
Measure	Black Participant-Concordant Sender-Expert Sender-Standard Signal Black participants who are assigned to a concordant-exper sender delivering a standard signal.	Black Participant-Concordant Sender-Lay Sender-Standard Signal Black participants who are assigned to a concordant-lay sender delivering a standard signal.	Black Participant-Discordant Sender-Expert Sender-Standard Signal Black participants who are assigned to a discordant-exper sender delivering a standard signal.	Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal Black participants who are assigned to a discordant-exper sender delivering a acknowledgement signal.	White Participant-Concordant Sender-Expert Sender-Standard Signal White participants who are assigned to a concordant-exper sender delivering a standard signal.	White Participant-Discordant Sender-Expert Sender-Standard Signal White participants who are assigned to a discordant-exper sender delivering a standard signal.
Overall Study STARTED	479	496	478	485	649	658
Overall Study COMPLETED	415	430	417	410	614	607
Overall Study NOT COMPLETED	64	66	61	75	35	51

Reasons for withdrawal

Withdrawal data not reported

Baseline Characteristics

Persuasion in Medicine: Experimental Evidence on Sender and Signal Effects

Baseline characteristics by cohort

Baseline characteristics by cohort
Measure	Black Participant-Concordant Sender-Expert Sender-Standard Signal n=415 Participants Black participants who are assigned to a concordant-exper sender delivering a standard signal.	Black Participant-Concordant Sender-Lay Sender-Standard Signal n=430 Participants Black participants who are assigned to a concordant-lay sender delivering a standard signal.	Black Participant-Discordant Sender-Expert Sender-Standard Signal n=417 Participants Black participants who are assigned to a discordant-exper sender delivering a standard signal.	Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal n=410 Participants Black participants who are assigned to a discordant-exper sender delivering a acknowledgement signal.	White Participant-Concordant Sender-Expert Sender-Standard Signal n=614 Participants White participants who are assigned to a concordant-exper sender delivering a standard signal.	White Participant-Discordant Sender-Expert Sender-Standard Signal n=607 Participants White participants who are assigned to a discordant-exper sender delivering a standard signal.	Total n=2893 Participants Total of all reporting groups
Age, Continuous	35.92048 years STANDARD_DEVIATION 6.717987 • n=5 Participants	35.53023 years STANDARD_DEVIATION 6.219959 • n=7 Participants	36.1247 years STANDARD_DEVIATION 6.46877 • n=5 Participants	35.90488 years STANDARD_DEVIATION 6.857133 • n=4 Participants	38.11401 years STANDARD_DEVIATION 6.773668 • n=21 Participants	38.16474 years STANDARD_DEVIATION 6.760583 • n=8 Participants	36.82613 years STANDARD_DEVIATION 6.742281 • n=8 Participants
Sex: Female, Male Female	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	0 Participants n=8 Participants
Sex: Female, Male Male	415 Participants n=5 Participants	430 Participants n=7 Participants	417 Participants n=5 Participants	410 Participants n=4 Participants	614 Participants n=21 Participants	607 Participants n=8 Participants	2893 Participants n=8 Participants
Race (NIH/OMB) American Indian or Alaska Native	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	0 Participants n=8 Participants
Race (NIH/OMB) Asian	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	0 Participants n=8 Participants
Race (NIH/OMB) Native Hawaiian or Other Pacific Islander	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	0 Participants n=8 Participants
Race (NIH/OMB) Black or African American	415 Participants n=5 Participants	430 Participants n=7 Participants	417 Participants n=5 Participants	410 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	1672 Participants n=8 Participants
Race (NIH/OMB) White	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	614 Participants n=21 Participants	607 Participants n=8 Participants	1221 Participants n=8 Participants
Race (NIH/OMB) More than one race	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	0 Participants n=8 Participants
Race (NIH/OMB) Unknown or Not Reported	0 Participants n=5 Participants	0 Participants n=7 Participants	0 Participants n=5 Participants	0 Participants n=4 Participants	0 Participants n=21 Participants	0 Participants n=8 Participants	0 Participants n=8 Participants
Region of Enrollment United States	415 participants n=5 Participants	430 participants n=7 Participants	417 participants n=5 Participants	410 participants n=4 Participants	614 participants n=21 Participants	607 participants n=8 Participants	2893 participants n=8 Participants
High School Completed	359 Participants n=5 Participants	380 Participants n=7 Participants	374 Participants n=5 Participants	355 Participants n=4 Participants	554 Participants n=21 Participants	533 Participants n=8 Participants	2555 Participants n=8 Participants
High School Not completed	56 Participants n=5 Participants	50 Participants n=7 Participants	43 Participants n=5 Participants	55 Participants n=4 Participants	60 Participants n=21 Participants	74 Participants n=8 Participants	338 Participants n=8 Participants
Married	71 Participants n=5 Participants	86 Participants n=7 Participants	89 Participants n=5 Participants	77 Participants n=4 Participants	202 Participants n=21 Participants	186 Participants n=8 Participants	711 Participants n=8 Participants
Insured	234 Participants n=5 Participants	251 Participants n=7 Participants	245 Participants n=5 Participants	239 Participants n=4 Participants	403 Participants n=21 Participants	391 Participants n=8 Participants	1763 Participants n=8 Participants
Subjective Health Status	3.522892 units on a scale STANDARD_DEVIATION 1.009056 • n=5 Participants	3.751163 units on a scale STANDARD_DEVIATION 1.003925 • n=7 Participants	3.642686 units on a scale STANDARD_DEVIATION 1.0069 • n=5 Participants	3.653659 units on a scale STANDARD_DEVIATION 1.056843 • n=4 Participants	3.21987 units on a scale STANDARD_DEVIATION .992071 • n=21 Participants	3.237232 units on a scale STANDARD_DEVIATION .9965605 • n=8 Participants	3.468372 units on a scale STANDARD_DEVIATION 1.030375 • n=8 Participants
Subjective Flu Shot Cost	43.91588 dollar STANDARD_DEVIATION 91.2157 • n=5 Participants	42.74393 dollar STANDARD_DEVIATION 85.77928 • n=7 Participants	38.14388 dollar STANDARD_DEVIATION 79.24301 • n=5 Participants	33.8441 dollar STANDARD_DEVIATION 72.69948 • n=4 Participants	25.28546 dollar STANDARD_DEVIATION 50.80593 • n=21 Participants	25.02221 dollar STANDARD_DEVIATION 48.44235 • n=8 Participants	33.56405 dollar STANDARD_DEVIATION 70.94346 • n=8 Participants
Has Primary Care Provider	189 Participants n=5 Participants	178 Participants n=7 Participants	192 Participants n=5 Participants	171 Participants n=4 Participants	321 Participants n=21 Participants	323 Participants n=8 Participants	1374 Participants n=8 Participants
(Prior) Flu Vaccine Intent	2.554217 units on a scale STANDARD_DEVIATION 3.243176 • n=5 Participants	2.786047 units on a scale STANDARD_DEVIATION 3.324808 • n=7 Participants	2.446043 units on a scale STANDARD_DEVIATION 3.307566 • n=5 Participants	2.485366 units on a scale STANDARD_DEVIATION 3.147713 • n=4 Participants	2.596091 units on a scale STANDARD_DEVIATION 3.146201 • n=21 Participants	2.52883 units on a scale STANDARD_DEVIATION 3.237376 • n=8 Participants	2.566886 units on a scale STANDARD_DEVIATION 3.228842 • n=8 Participants
(Prior) Likelihood of Contracting Flu	2.342169 units on a scale STANDARD_DEVIATION 2.853683 • n=5 Participants	2.072093 units on a scale STANDARD_DEVIATION 2.791414 • n=7 Participants	2.143885 units on a scale STANDARD_DEVIATION 2.823905 • n=5 Participants	2.304878 units on a scale STANDARD_DEVIATION 2.86659 • n=4 Participants	2.760586 units on a scale STANDARD_DEVIATION 2.603112 • n=21 Participants	2.912685 units on a scale STANDARD_DEVIATION 2.701415 • n=8 Participants	2.476668 units on a scale STANDARD_DEVIATION 2.774967 • n=8 Participants

PRIMARY outcome

Timeframe: This outcome was assessed during Baseline survey, which took approximately 20 minutes.

Population: Arms/Groups are combined as described in each arm/group description in order to test hypotheses prespecified in the analysis plan as well as to measure the effect of the given intervention on the outcome among the group of participants in Arm A compared to Arm B. The intervention and Arms A and B are described in each arm/group description.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=832 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=430 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=827 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=845 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
Rating of Sender	.1828917 index Interval 0.0506082 to 0.3151752	-.0750636 index Interval -0.1870341 to 0.0369069	.0997998 index Interval -0.0345829 to 0.2341826	-.5397662 index Interval -0.6789624 to -0.4005701

PRIMARY outcome

Timeframe: This outcome was assessed during Baseline survey, which took approximately 20 minutes.

This is an inverse-covariance-weighted index comprised of responses to survey questions on: 1. recommending the video to friends and family measured on a scale of 0 to 10; 2. recommending the flu shot to friends and family measured on a scale of 0 to 10; 3. and the respondent's assessment of the extent to which the information contained in the video was useful measured on a scale of 1 to 5. It is standardized to a mean of 0 and standard deviation of 1. An index score of 0 is equal to the mean. Scores less than 0 indicate lower ratings of the signal rated by participants in Arm A (described in each arm/group description) compared to the rating rated by participants in Arm B (also described in each arm/group description). Scores greater than 0 indicate higher ratings of the signal by Arm A compared to by Arm B. For example, an index score of 0.14 indicates that participants in Arm A rate the signal 0.14 index (in standard deviation units) higher than those in Arm B.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=832 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=1221 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=827 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=845 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
Rating of Signal	.1387205 index Interval 0.0003558 to 0.2770851	-.0088105 index Interval -0.1199617 to 0.1023406	.1423725 index Interval 0.0062946 to 0.2784504	-.0806321 index Interval -0.2126516 to 0.0513873

PRIMARY outcome

Timeframe: This outcome will be assessed during Baseline survey, which takes approximately 20 minutes.

This is an inverse-covariance-weighted index comprised of responses to survey questions on: 1. whether age group for whom the flu vaccine is recommended (measured as a binary outcome); and 2. whether the flu shot contains the flu virus (recall of information discussed in the video) (measured as a binary outcome). It is standardized to a mean of 0 and standard deviation of 1. An index score of 0 is equal to the mean. Scores less than 0 indicate lower level of recalling signal content among participants in Arm A (described in each arm/group description) compared to the level among participants in Arm B (also described in each arm/group description). Scores greater than 0 indicate a higher level of recalling signal content among participants in Arm A compared to those in Arm B. For example, an index score of 0.02 indicates that participants in Arm A recall signal content 0.02 index (in standard deviation units) higher than those in Arm B.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=832 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=1221 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=827 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=845 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
Signal Content Recall	-.0062555 index Interval -0.1415234 to 0.1290124	.0193832 index Interval -0.0926014 to 0.1313679	.0041984 index Interval -0.131372 to 0.1397687	.1166304 index Interval -0.0149457 to 0.2482066

PRIMARY outcome

Timeframe: This outcome will be assessed during Baseline survey, which takes approximately 20 minutes.

This is an inverse-covariance-weighted index comprised of: 1. the point belief on the likelihood to contract the flu from the flu shot measured by a Likert scale and rescaled to have a support of -1 to 1. 2. the certainty measured with a balls and bins method and rescaled to have a support of -1 to 1. It is standardized to a mean of 0 and standard deviation of 1. An index score of 0 is equal to the mean. Scores less than 0 indicate a lower level of beliefs in the safety of flu vaccine among participants in Arm A (described in each arm/group description) compared to the level among participants in Arm B (also described in each arm/group description). Scores greater than 0 indicate a higher level of beliefs among participants in Arm A compared to those in Arm B. For example, an index score of -0.1 indicates that participants in Arm A hold such beliefs 0.1 index (in standard deviation units) lower than those in Arm B.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=832 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=1221 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=827 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=845 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
Safety Beliefs	-.0979564 index Interval -0.2331252 to 0.0372124	-.027714 index Interval -0.1410194 to 0.0855914	-.1065867 index Interval -0.2425802 to 0.0294069	-.0242752 index Interval -0.1580924 to 0.109542

PRIMARY outcome

Timeframe: This outcome was assessed during Baseline survey, which took approximately 20 minutes.

This is an inverse-covariance-weighted index comprised of revealed preference measures of demand for a free flu shot coupon: 1. Willingness to pay (WTP) for the coupon measured as $0, $1, $2, $5, or $10 2. demand for information regarding locations to redeem the coupon measured as a binary outcome It is standardized to a mean of 0 and standard deviation of 1. An index score of 0 is equal to the mean. Scores less than 0 indicate a lower level of interest in a free flu shot coupon distributed after survey completion among participants in Arm A (described in each arm/group description) compared to the level among participants in Arm B (also described in each arm/group description). Scores greater than 0 indicate a higher level of interest among participants in Arm A compared to those in Arm B. For example, an index score of -0.01 indicates that participants in Arm A hold such interest -0.01 index (in standard deviation units) higher than those in Arm B.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=832 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=1221 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=827 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=845 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
Coupon Interest	-.0078177 index Interval -0.1389825 to 0.123347	-.0827849 index Interval -0.1924382 to 0.0268684	.0280955 index Interval -0.1069421 to 0.1631332	-.0162865 index Interval -0.1513941 to 0.1188211

PRIMARY outcome

Timeframe: This outcome was assessed after the video treatment during Baseline survey, which took approximately 20 minutes.

This is the respondent's likelihood to receive the flu vaccine before the end of the flu season. This was elicited on an 11-point Likert scale (0 extremely unwilling, 10 extremely willing) after the video message treatment. We re-scaled this outcome to have a scale of 0 to 1. The reported results of the outcome compare participants in Arm A and Arm B (described in each arm/group description). For example, an intent of 0.03 indicates that participants in Arm A are 0.03 units on a scale more likely to intent to take up the flu vaccine compared to those in Arm B.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=832 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=1221 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=827 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=845 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
(Posterior) Flu Vaccine Intent	.0263281 units on a scale Interval -0.0236633 to 0.0763195	.0034755 units on a scale Interval -0.037638 to 0.044589	.0268659 units on a scale Interval -0.0226463 to 0.0763781	.0187403 units on a scale Interval -0.0305011 to 0.0679817

PRIMARY outcome

Timeframe: This outcome was assessed after the video treatment during Baseline survey, which took approximately 20 minutes.

This is the respondent's likelihood to take up the COVID-19 vaccine if made available free of charge. This was elicited on an 11-point Likert scale (0 extremely unwilling, 10 extremely willing) after the video message treatment. We re-scaled this outcome to have a scale of 0 to 1. The reported results of the outcome compare participants in Arm A and Arm B (described in each arm/group description). For example, an intent of 0.04 indicates that participants in Arm A are 0.04 units on a scale more likely to intent to take up the COVID-19 vaccine compared to those in Arm B.

Outcome measures

Outcome measures
Measure	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - Black Participants n=587 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Concordant (Arm A) vs. Discordant Expert Sender (Arm B) - White Participants n=866 Participants This group combines the following arms to measure the effect of racial concordance between sender and receiver (i.e. intervention) among White Participants: 1. White Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm A) 2. White Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 1 prespecified in the analysis plan. The hypothesis tests how racial concordance between an expert sender and receiver affects the outcome among participants assigned to Arm A compared to those who are assigned to Arm B.	Acknowledgement (Arm A) vs. Standard Signal (Arm B) - Black Participants n=581 Participants This group combines the following arms to measure the effect of an acknowledgement signal (i.e. intervention) among Black Participants: 1. Black Participant-Discordant Sender-Expert Sender-Acknowledgement Signal (i.e. Arm A) 2. Black Participant-Discordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 4 prespecified in the analysis plan. The hypothesis tests how a signal that acknowledges past injustices affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.	Layperson Sender (Arm A) vs. Expert Sender (Arm B) - Black Participants n=592 Participants This group combines the following arms to measure the effect of a layperson sender (i.e. intervention) among Black Participants: 1. Black Participant-Concordant Sender-Lay Sender-Standard Signal (i.e. Arm A) 2. Black Participant-Concordant Sender-Expert Sender-Standard Signal (i.e. Arm B) The arms are combined to test Hypothesis 3 prespecified in the analysis plan. The hypothesis tests how a layperson sender affects the outcome among participants assigned to Arm A compared to those assigned to Arm B.
COVID-19 Vaccine Intent	.0350147 units on a scale Interval -0.0251663 to 0.0951958	.009072 units on a scale Interval -0.0404259 to 0.0585699	.0544277 units on a scale Interval -0.0064546 to 0.11531	.0881949 units on a scale Interval 0.0291914 to 0.1471983

PRIMARY outcome

Timeframe: This outcome was assessed during the follow-up survey which was conducted at least two weeks after the initial baseline survey.

This is respondents' self-report in the follow-up survey on whether they or their family members had received the flu shot or redeemed the coupon since the intervention. The reported results of the outcome compare participants in Arm A and Arm B (described in each arm/group description). For example, 15.0 percentage points indicate that participants in Arm A are more likely to report that they and/or another household member received the flu vaccine by 15.p percentage points compared to those in Arm B.